MV> Wow, that is a lot sooner than expected, considering that Geoff Huston MV> estimated this to happen on 8 april 2015: MV> http://www.potaroo.net/tools/ipv4/
Ah, now see where the discrepancy comes from:
> Notes:
> "Exhaustion" is defined here as the time when the pool of available > addresses in each RIR reaches the threshold of no more general use > allocations of IPv4 addresses. As ARIN have already reserved a /10 for > the transition to Ipv6 policy, the low point for ARIN is a completely > depleted general use pool. For AFRINIC and LACNIC it the threshold is a > total of a /11 remaining in their available address pool. This > calculation also takes into account the redistribution of the IANA > Global Address pool, and in the simulation of exhaustion these addresses > are redistributed to the RIRs according to the policy.
So Huston uses a different definition for "exhaustion" for the different RIR's. According to Huston ARIN is not exhausted yet, that will happen when all of the last /8 is also gone. Which he expects to last another year.
APNIC and RIPE were considered "exhausted" when they reached their last /8.